Introduction

Cache Rocket keeps your website and API caches warm. It visits your URLs on a schedule, the same way a real visitor would, so the expensive work of building a page has already happened before anyone asks for it.

You do not install anything on your server to get started. You verify that you own a domain, create a warmer, and Cache Rocket does the rest over ordinary HTTP requests.

On paid plans, Cache Rocket also offers [Managed CDN](/documentation/managed-cdn) and cloud Media optimization—image formats, Critical CSS, LQIP, and PageSpeed Insights—primarily through the WordPress plugin, alongside CDN integrations for purge → rewarm on your existing edge.

The problem: the first visitor pays#

Almost every fast website is fast because something is cached — a CDN edge node, a reverse proxy like nginx or Varnish, a WordPress page cache, or your framework's own render cache. Caches are wonderful, but they all expire.

When a cached copy expires, it is gone. The next person who asks for that page does not get the fast cached version. Their request falls through to your origin server, which has to rebuild the page from scratch: run the code, query the database, render the template, and send it back. That person waits. We call this a cold page, and the delay is the *cold-start penalty*.

It is worse than it sounds, because cold requests land on exactly the people you care about most:

• The reader who clicks your newly published article thirty seconds after it goes live.
• The shopper arriving from an ad campaign on a landing page nobody has touched all night.
• Everyone at once, right after you deploy and every cache is emptied simultaneously.
• Visitors in a region whose CDN edge node has never served that page before.

The fix: warm it first#

Cache Rocket removes the cold-start penalty by being the first visitor itself. It requests your URLs before real people do, which forces every cache layer in front of your origin to store a fresh copy. By the time a human arrives, the page is already in cache and is served instantly.

That is the entire idea. Everything else exists to make it practical: deciding *which* URLs, *how often*, *from where*, *how fast*, and proving it actually worked.

What you can warm#

Is Cache Rocket right for you?#

It helps most when pages are expensive to build and traffic is uneven enough that caches go cold between visitors.

It helps less if your pages are already static files served from disk with no expiry, or if every response is personalised per user and therefore never cacheable in the first place.

How to use these docs#

The pages are written to be read in order, but each one stands alone. If you are brand new, this is the fastest route:

1. 1

   Read How it works

   Ten minutes on what actually happens during a warm run. Worth it before you start changing settings.

2. 2

   Follow the Quickstart

   A guided first setup: create an account, verify a hostname, build your first warmer, and confirm it worked.

3. 3

   Skim Core concepts

   The vocabulary — entry URLs, depth, intervals, warmth score. The rest of the docs assume these words.

4. 4

   Then go deeper where you need to

   The sidebar is grouped by task: Setup, Warming, Monitoring, Integrations, Account, and Help.